The operation process of an intelligent service requires support of service voice announcement. The voice announcement can be used specially by the services and the users, or can be used commonly by several services. Various kinds of selection and operation can be performed by the users according to the prompts of the service announcement. The voice announcement of the service is one kind of the special resources of SRF (Specialized Resource Function), while SRF also provides the management to the resources. Service Management Function (SMF) manages SRF, and SMF only operates under the request of other functional entities in any case, rather than activating by itself. Today, because the standard interface between SRF and SMF is not existed yet, an essential management relation performed by SMF for SRF is existed. Owing to the comprehensiveness of SRF, operation, management, and maintenance of SRF will be a part of operation, management, and maintenance of an entire switching system, and it can be realized by various ways, such as operator commands, as follows:                (1) adding or deleting record announcement of existing services;        (2) generating the record announcement or the words by a voice edit system, which is an off-line tool separated entirely from SRF;        (3) increasing a voice element without affecting a call processing ability of a switch system, so long as the remaining capacity is existed in a SRF memory.        (4) integrating a method of SRF hardware expanding and an existing method in the switch system.        
Today, the voice announcement of the intelligent service can be stored in an individual IP (Intelligent Peripheral) or an intra-IP of an integrated SSP (Service Switching Point). There are many loading methods, but all of these methods require a lot of interposing and manual operations carried out by a user. Loading of a service voice must be carried out in the case that the service quantity is the minimum, as well as an occupation rate of system resources is the smallest at this time. However, since the location of SSP or IP in the network may be at an end office, or may be at a tandem office, a tall office, or an international exchange, when a new service is opened, voice content in a voice single board at different places must be updated simultaneously to ensure voice versions on all of the SSP or IP modules are identical, this requires a large amount of staffs and time. Changing or updating the single boards or chips, which store voice information, are necessary in the existing loading methods of the user voice announcement. Thus, the normal operation of the service will be affected during a voice loading process, thereby increasing the operation cost of the service.
During an operation process of an existing intelligent network or other intelligent platform, a voice of many telephone-card services does not change frequently, so that the loading of the voice can be completed at the beginning of opening of the service. However, when a service (such as VOT (telephone voting service), MAS (mass call service), AD (telephone advertising service), etc.) associated with a user sponsored service is to be opened, a service voice associated with the user sponsored service must be configured uniformly into the related individual IP, in the network, or in the intra-IP of the integrated SSP, to ensure that the service is opened successfully. Therefore, popularizing to use the service will be affected by a method of loading a user voice announcement into the individual IP, in the network, or in the intra-IP of the integrated SSP.
Changing or updating a voice single board or chip which stores the voice information is typically involved in the existing loading methods of the user voice announcement. Accordingly, the normal operation of the service is affected during the voice loading process, thereby increasing the operation cost of the service.
With the rapid development of an intelligent network service, the demand for service changes or new services has been increasing significantly. However, the rapid development of new services and changes of service are limited by the existing service voice loading methods. For example, in an AD (Advertisement) service, a user of the service may require to modify the service voice at any time based on the service feature of the AD service. If adding or deleting a voice announcement is implemented by using the existing method that manages the user voice announcement on a switch, the difficulty to open such a service on a larger scale network is noticeable. In the meantime, it is extremely difficult to manage those voice announcements on a switch system while ensuring the consistence of the voice version in the respective SSP module.
Accordingly, there is a need for a method and network architecture to overcome the limitation of the existing service voice loading method.